arXiv subjects

Said Safi

Publications and source records attributed to Said Safi.

3 recordsLinked to original sources

arXiv2016

Facial Surface Analysis using Iso-Geodesic Curves in Three Dimensional Face Recognition System

In this paper, we present an automatic 3D face recognition system. This system is based on the representation of human faces surfaces as collections of Iso-Geodesic Curves (IGC) using 3D Fast Marching algorithm. To compare two facial surfaces, we compute a geodesic distance between a pair of facial curves using a Riemannian geometry. In the classifying step, we use: Neural Networks (NN), K-Nearest Neighbor (KNN) and Support Vector Machines (SVM). To test this method and evaluate its performance, a simulation series of experiments were performed on 3D Shape REtrieval Contest 2008 database (SHREC2008).

cs.CV↗

arXiv2015

An original Propagator for large array

In this paper, we demonstrate that when the ratio $n$ of the number of antenna elements $N$ to the number $P$ of radiating sources is superior or equal to $2$, then it is possible to choose a propagator from a set of $n(n+1)/2-1$ operators to compute the Angles of Arrival (AoA) of the narrowband incoming waves. This new non eigenbased approach is efficient when the Signal to Noise Ratio (SNR) is moderate, and gives multitude of possibilities, that are dependent of the random data, to construct the complex sets whose columns are orthogonal to the signal subspace generated by the radiating sources. Elementary examples are given for $n=3$, $n=4$ and $n=6$. The simulation results are presented to illustrate the performance of the proposed computational methods.

cs.IT↗

arXiv2014

Angle of Arrival Detection with Fifth Order Phase Operators

In this paper, a fifth order propagator operators are proposed for estimating the Angles Of Arrival (AOA) of narrowband electromagnetic waves impinging on antenna array when its number of sensors is larger than the number of radiating sources. The array response matrix is partitioned into five linearly dependent phases to construct the noise projector using five different propagators from non diagonal blocks of the spectral matrice of the received data; hence, five different estimators are proposed to estimate the angles of the sources. The simulation results proved the performance of the proposed estimators in the presence of white noise comparatively to high resolution eigen based spectra.

cs.IT↗